US 12,323,288 B2
Method and device for determining a control signal
Duc Tam Nguyen, Karlsruhe (DE); Michael Klar, Bad Friedrichshall (DE); and Zhongyu Lou, Renningen (DE)
Assigned to ROBERT BOSCH GMBH, Stuttgart (DE)
Appl. No. 17/262,181
Filed by Robert Bosch GmbH, Stuttgart (DE)
PCT Filed Aug. 14, 2019, PCT No. PCT/EP2019/071802
§ 371(c)(1), (2) Date Jan. 21, 2021,
PCT Pub. No. WO2020/064211, PCT Pub. Date Apr. 2, 2020.
Claims priority of application No. 102018216609.9 (DE), filed on Sep. 27, 2018; and application No. 102018218834.3 (DE), filed on Nov. 5, 2018.
Prior Publication US 2021/0294278 A1, Sep. 23, 2021
Int. Cl. H04L 41/0631 (2022.01); G05B 13/02 (2006.01); G06N 3/08 (2023.01)
CPC H04L 41/0631 (2013.01) [G05B 13/027 (2013.01); G06N 3/08 (2013.01)] 17 Claims
OG exemplary drawing
 
1. A computer-implemented method for classifying an input signal that is ascertained as a function of an output signal of a sensor as to whether or not the input signal has an anomaly, the method comprising:
deciding, as a function of an output signal of an autoencoder to which the input signal is supplied, whether or not the input signal has the anomaly;
wherein:
the autoencoder includes an encoder and a plurality of decoders;
the encoder ascertains an intermediate variable as a function of the input signal;
the plurality of decoders ascertain the output signal as a function of the intermediate variable;
the autoencoder ascertains, for the input signal supplied to the autoencoder, a plurality of hypotheses for reconstruction of the input signal;
each of the plurality of decoders provides a respective one of a plurality of output variables, with a respective one of the plurality of hypotheses being respectively provided to a respective one of the decoders as a function of the output variables; and
the output signal is ascertained as a function of the plurality of hypotheses.